53 currently known mutations in the human microtubule associated protein tau gene (MAPT) cause frontotemporal dementia, characterized by behavioral disturbances, cognitive impairment and parkinsonism. These mutations have been described as frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17). The dominant mutations affect men and woman equally and have an average onset of 49 years of age. We recently reported that human MAPT generates two human-specific circular RNAs, caused by back-splicing of exon 12 to either exon 7 or 10. Intriguingly, 46 of the 53 known FTDP-17 mutations are located in these circular RNAs. The FTDP-17 mutations V337M and K317M introduce a start codon in tau 12->10 circRNA, which lacks a start codon in all reading frames. Highly unexpectedly, transfection experiments with reporter genes showed that the circular RNAs are translated in HEK293 cells. This suggests that the V337M and K317M mutations act by generating a protein from the tau 12->10 circular RNA. The 12->10 circRNA contains 288 nt and no stop codon. Since 288 nt/3 = 96 amino acids, the presence of V337M and K317M leads to a rolling circle translation creating tau polymers. We observed one to three rounds of translation, i.e. a protein predicted of 96-288 amino acids in both mutants. The tau circ12->7 RNA is weakly expressed but has its own start codon and again we observe translation using reporter gene systems. We postulate that other FTDP-17 mutations influence translatability, stability and/or location of the tau circ12->7 RNA and its protein products. We will test the hypothesis that FTDP-17 mutations act through tau circular RNAs by generating peptides that promote the formation of neurofibrillary tangles (NFT) in two specific aims. First, we will characterize the influence of FTDP-17 mutation on translatability of the tau 12->10 and 12->7 circRNAs, investigate the translational initiation and possible contributions to NFT formation. In the second Aim, we will Investigate the influence of tau circular RNAs harboring FTDP-17 mutations on neurons and zebrafish and explore treatment options. These experiments are highly innovative as they are one of the first demonstration of a translation of a natural circular RNA in a rolling circle mechanism. The translation is completely unexpected, as the circular RNAs lack a trimethyl guanosine cap and a known ribosomal entry site and thus according to current `textbook knowledge, should not be translated. The findings are significant, as they could provide a new model for the molecular mechanism that generates tau aggregates in FTDP-17. Although the FTDP-17 mutations are rare, they can be considered `experiments of nature', which will help understand the more frequent tau pathology in Alzheimer's disease.